This paper proposes wall function models to simulate the heat transfer around a cylinder in cross flow with an isothermal and rough surface. The selected case has similitudes with aircraft wing icing: the ice roughness shape, height and distribution. Moreover, the flow is somewhat similar to that found on iced airfoil; and the surface is isothermal like when icing. The Reynolds-Averaged Navier-Stokes, turbulence, energy and mass conservation 7-equation system is solved by two Computational Fluid Dynamics (CFD) codes. To represent accurately the effects of roughness on the heat transfer, the present authors had to modify both codes and to propose new thermal wall functions for them. In addition, it was implemented a momentum wall function that is not so common in CFD codes but it is a standard in aircraft icing simulation. Basically, the present paper is focused on presenting the simulation results improvement obtained by the implementation of thermal wall functions that also considered the effect thermal resistance of viscous sub-layer on convective heat transfer. The numerical results were compared to experimental data of heat transfer around cylinders with an isothermally heated surface and equivalent sand grain roughness height of K s /D = 900 • 10 −5. For the selected case, the flow regime was trans-critical at Reynolds numbers Re = u e • D/ν = 2.2 • 10 5. Due to significant blockage effects present in the experimental data, the tunnel walls were also meshed and simulated. In sum, the implementation into CFD codes was considered adequate because the results were close to experimental data around the whole cylinder surface, not only along the leading edge or the separated region. The results accuracy were improved when compared with CFD factory original models. However, the results indicate that further works on wall functions and their validation are required before using CFD in wing aircraft icing.

The study of flame stability is very important to the design of burners used in industrial ovens and furnaces, combustion chambers of gas turbines and flares; and fuel substitution in burners. There is a range of conditions (for example gas velocity at the nozzle exit and jet fuel concentration in the gas mixture) at which stable combustion can be maintained, being limited by two phenomena called liftoff and blowout. Lift-off is the detachment of the flame from the fuel nozzle, and blowout its detachment and extinction. Operating conditions close to stability limits should be avoided for security reasons. Many theories have been published to describe the blowout and lifted characteristics of turbulent jet diffusion flames. This document presents some theories, as well as the assumptions and physical processes considered responsible for these phenomena (liftoff and blowout). Correlations for predicting the blowout velocity and experimental results available in the literature are also shown. A new correlation is proposed, which is based on large-scale motions observed in turbulent jets and the dimensionless Damköhler number (ratio of the characteristic chemical reaction time and the time associated with the mixing of reentrained hot products into fresh reactants). Comparisons between the predictions of the proposed correlation with experimental results and predictions of other correlations available in the literature were performed for different fuels and nozzle diameters. The proposed correlation showed good agreement with the experimental results. The analyses developed in this work allow us to conclude that the blowout velocity of the turbulent diffusion flame depends on the fuel properties, characteristics of the nozzle, the environmental conditions and the Damköhler number.